CSP implementation for MCAP

CSP implementation for MCAP

[Elvis Pfützenreuter]

This is the repository for the CSP implementation, rebased over the recently accepted MCAP, for your appreciation:

git://gitorious.org/bluez-epx/bluez-epx.git csp

or

http://www.gitorious.org/bluez-epx/bluez-epx/commits/csp

Re: CSP implementation for MCAP

[Johan Hedberg]

Hi Elvis,

On Thu, Sep 02, 2010, Elvis Pfützenreuter wrote:
> This is the repository for the CSP implementation, rebased over the
> recently accepted MCAP, for your appreciation:
> 
> git://gitorious.org/bluez-epx/bluez-epx.git csp

A couple of quick comments:

- Several of the commits in the tree have the author as
  "epx <epx@xenicall.(none)>". That needs fixing.

- I'm not so happy about all the ifdefs in the code and the way that
  they are used. So this may need some rethinking. Probably we could
  have the code always compiled in and have a runtime variable to
  disable its use.

Johan

Re: CSP implementation for MCAP

[Elvis Pfützenreuter]

On Sep 2, 2010, at 5:28 PM, Johan Hedberg wrote:

> Hi Elvis,
> 
> On Thu, Sep 02, 2010, Elvis Pfützenreuter wrote:
>> This is the repository for the CSP implementation, rebased over the
>> recently accepted MCAP, for your appreciation:
>> 
>> git://gitorious.org/bluez-epx/bluez-epx.git csp
> 
> A couple of quick comments:
> 
> - Several of the commits in the tree have the author as
>  "epx <epx@xenicall.(none)>". That needs fixing.

Ouch :(

> - I'm not so happy about all the ifdefs in the code and the way that
>  they are used. So this may need some rethinking. Probably we could
>  have the code always compiled in and have a runtime variable to
>  disable its use.

Such variable already exists, and starts turned off. So could I remove the --enable-mcap-csp flag altogether?

> Johan

Re: CSP implementation for MCAP

[Johan Hedberg]

Hi Elvis,

On Thu, Sep 02, 2010, Elvis Pfützenreuter wrote:
> > - I'm not so happy about all the ifdefs in the code and the way that
> >  they are used. So this may need some rethinking. Probably we could
> >  have the code always compiled in and have a runtime variable to
> >  disable its use.
> 
> Such variable already exists, and starts turned off. So could I remove
> the --enable-mcap-csp flag altogether?

Yes, I think a single --enable-mcap that covers also csp should be
enough.

Johan

Re: CSP implementation for MCAP

[Elvis Pfützenreuter]

On Sep 2, 2010, at 5:40 PM, Johan Hedberg wrote:

> Hi Elvis,
> 
> On Thu, Sep 02, 2010, Elvis Pfützenreuter wrote:
>>> - I'm not so happy about all the ifdefs in the code and the way that
>>> they are used. So this may need some rethinking. Probably we could
>>> have the code always compiled in and have a runtime variable to
>>> disable its use.
>> 
>> Such variable already exists, and starts turned off. So could I remove
>> the --enable-mcap-csp flag altogether?
> 
> Yes, I think a single --enable-mcap that covers also csp should be
> enough.

Ok, those first changes have been made and tested. Waiting for the next round.

Re: CSP implementation for MCAP

[Santiago Carot-Nemesio]

Hi Elvis,

On 09/02/10 20:58, Elvis Pfützenreuter wrote:
> This is the repository for the CSP implementation, rebased over the recently accepted MCAP, for your appreciation:
>
> git://gitorious.org/bluez-epx/bluez-epx.git csp
>

I have been divin in CSP code and at first look I found some issues that 
I would like discuss with you.

The call to function mcap_sync_stop(mcl) it's only neccesary in static 
close_mcl function. You can remove the second call in line 799 when user 
closes explicitly the mcl due that first sync_stop will be called from 
watcher set in control channel when its socket is closed.

Second thing is related to code structure. Because standard op codes and 
close synchronization protocol have separate logic, it may better put 
csp parameters away from mcl in a separate structure to avoid mess all 
protocol logic in a big mcl structure, I was thinking in something like 
this:

struct mcap_mcl {
	/* Op code parameters */
	....
	struct	mcap_csp	*csp;
}

struct mcap_csp {
	uint64_t	base_tmstamp;
	struct timespec	base_time;
	guint		local_caps;
	guint		remote_caps;
	guint		rem_req_acc;
	guint		ind_expected;
	MCAPCtrl	csp_req;
	guint		ind_timer;
	guint		set_timer;
	void		*set_data;
	gint		dev_id;
	gint		dev_hci_fd;
	void		*csp_priv_data;
};

Because CSP is optional we can reserve memory for csp only when we will 
use Clock Synchronization protocol.
Of course, I know that it depends on invidivual taste, but I would like 
to comment this issue.

Comments are welcome.

Regards.

Re: CSP implementation for MCAP

[Elvis Pfützenreuter]

On Sep 3, 2010, at 5:03 AM, Santiago Carot-Nemesio wrote:

> Hi Elvis,
> 
> On 09/02/10 20:58, Elvis Pfützenreuter wrote:
>> This is the repository for the CSP implementation, rebased over the recently accepted MCAP, for your appreciation:
>> 
>> git://gitorious.org/bluez-epx/bluez-epx.git csp
>> 
> 
> I have been divin in CSP code and at first look I found some issues that I would like discuss with you.
> 
> The call to function mcap_sync_stop(mcl) it's only neccesary in static close_mcl function. You can remove the second call in line 799 when user closes explicitly the mcl due that first sync_stop will be called from watcher set in control channel when its socket is closed.
> 
> Second thing is related to code structure. Because standard op codes and close synchronization protocol have separate logic, it may better put csp parameters away from mcl in a separate structure to avoid mess all protocol logic in a big mcl structure, I was thinking in something like this:
> 
> struct mcap_mcl {
> 	/* Op code parameters */
> 	....
> 	struct	mcap_csp	*csp;
> }
> 
> struct mcap_csp {
> 	uint64_t	base_tmstamp;
> 	struct timespec	base_time;
> 	guint		local_caps;
> 	guint		remote_caps;
> 	guint		rem_req_acc;
> 	guint		ind_expected;
> 	MCAPCtrl	csp_req;
> 	guint		ind_timer;
> 	guint		set_timer;
> 	void		*set_data;
> 	gint		dev_id;
> 	gint		dev_hci_fd;
> 	void		*csp_priv_data;
> };
> 
> Because CSP is optional we can reserve memory for csp only when we will use Clock Synchronization protocol.
> Of course, I know that it depends on invidivual taste, but I would like to comment this issue.

Sounds sensible.

> 
> Comments are welcome.
> 
> Regards.
> 
> 

Re: CSP implementation for MCAP

[Elvis Pfützenreuter]

On 02/09/2010, at 15:58, Elvis Pfützenreuter wrote:

> This is the repository for the CSP implementation, rebased over the recently accepted MCAP, for your appreciation:
> 
> git://gitorious.org/bluez-epx/bluez-epx.git csp
> 
> or
> 
> http://www.gitorious.org/bluez-epx/bluez-epx/commits/csp--

And I am happy to announce that this CSP implementation passed on PTS :) Fixes are on topmost patch of the repository.

Re: CSP implementation for MCAP

[Johan Hedberg]

Hi Elvis,

On Wed, Sep 15, 2010, Elvis Pfützenreuter wrote:
> > This is the repository for the CSP implementation, rebased over the
> > recently accepted MCAP, for your appreciation:
> > 
> > git://gitorious.org/bluez-epx/bluez-epx.git csp
> > 
> > or
> > 
> > http://www.gitorious.org/bluez-epx/bluez-epx/commits/csp--
> 
> And I am happy to announce that this CSP implementation passed on PTS
> :) Fixes are on topmost patch of the repository.

Good to hear about the progress with the PTS :)

Unfortunately you'll need to rebase again since I pushed some cleanups
to the MCAP code. Could you also get rid of the unnecessary
double-pointers and type casts which aren't needed anymore now that
mcap_send_data accepts void *.

Johan

Re: CSP implementation for MCAP

[José Antonio Santos Cadenas]

El Thursday 16 September 2010 14:39:00 Johan Hedberg escribió:
> Hi Elvis,
> 
> On Wed, Sep 15, 2010, Elvis Pfützenreuter wrote:
> > > This is the repository for the CSP implementation, rebased over the
> > > recently accepted MCAP, for your appreciation:
> > > 
> > > git://gitorious.org/bluez-epx/bluez-epx.git csp
> > > 
> > > or
> > > 
> > > http://www.gitorious.org/bluez-epx/bluez-epx/commits/csp--
> > 
> > And I am happy to announce that this CSP implementation passed on PTS
> > 
> > :) Fixes are on topmost patch of the repository.
> 
> Good to hear about the progress with the PTS :)
> 
> Unfortunately you'll need to rebase again since I pushed some cleanups
> to the MCAP code. Could you also get rid of the unnecessary
> double-pointers and type casts which aren't needed anymore now that
> mcap_send_data accepts void *.

There are also a lot of (! condition) with an extra space between the ! and 
the condition variable.

Also some use of g_malloc0 instead of g_new0. About this issue, remember that 
g_new0 initializes all the structure with 0 so the initialization don't need 
to do for some fields in mcap_sync_init function.

Regards.

> 
> Johan
> --
> To unsubscribe from this list: send the line "unsubscribe linux-bluetooth"
> in the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: CSP implementation for MCAP

[Elvis Pfützenreuter]

On 16/09/2010, at 09:39, Johan Hedberg wrote:

> Hi Elvis,
> 
> On Wed, Sep 15, 2010, Elvis Pfützenreuter wrote:
>>> This is the repository for the CSP implementation, rebased over the
>>> recently accepted MCAP, for your appreciation:
>>> 
>>> git://gitorious.org/bluez-epx/bluez-epx.git csp
>>> 
>>> or
>>> 
>>> http://www.gitorious.org/bluez-epx/bluez-epx/commits/csp--
>> 
>> And I am happy to announce that this CSP implementation passed on PTS
>> :) Fixes are on topmost patch of the repository.
> 
> Good to hear about the progress with the PTS :)
> 
> Unfortunately you'll need to rebase again since I pushed some cleanups
> to the MCAP code. Could you also get rid of the unnecessary
> double-pointers and type casts which aren't needed anymore now that
> mcap_send_data accepts void *.

It's done, and tested.

Re: CSP implementation for MCAP

[Johan Hedberg]

Hi Elvis,

On Thu, Sep 16, 2010, Elvis Pfützenreuter wrote:
> > On Wed, Sep 15, 2010, Elvis Pfützenreuter wrote:
> >>> This is the repository for the CSP implementation, rebased over the
> >>> recently accepted MCAP, for your appreciation:
> >>> 
> >>> git://gitorious.org/bluez-epx/bluez-epx.git csp
> >>> 
> >>> or
> >>> 
> >>> http://www.gitorious.org/bluez-epx/bluez-epx/commits/csp--
> >> 
> >> And I am happy to announce that this CSP implementation passed on PTS
> >> :) Fixes are on topmost patch of the repository.
> > 
> > Good to hear about the progress with the PTS :)
> > 
> > Unfortunately you'll need to rebase again since I pushed some cleanups
> > to the MCAP code. Could you also get rid of the unnecessary
> > double-pointers and type casts which aren't needed anymore now that
> > mcap_send_data accepts void *.
> 
> It's done, and tested.

Thanks. I still had to do quite a lot of coding style fixes as well as
plug a memory leak, but the patches are now pushed upstream.

Next steps:

- Get rid of the forward declarations of static functions
  wherever possible. It looked like you could accomplish that
  with a simple reordering of the functions.

- Get rid of raw HCI access (hci_open_dev). bluetoothd shouldn't use
  this type of sockets in the future at all. You'll need to add a proper
  callback to the adapter_opts and a function to plugins/hciopts.c and
  use that instead. Then export a function like btd_adapter_read_clock
  for the plugin to use.

Johan